Biosynthesis of Osmoregulated Periplasmic Glucans in Escherichia coli: The Phosphoethanolamine Transferase Is Encoded by opgE
Osmoregulated periplasmic glucans (OPGs) are oligosaccharides found in the periplasm of many Gram-negative bacteria. Glucose is the sole constitutive sugar and this backbone may be substituted by various kinds of molecules depending on the species. In E. coli, OPG are substituted by phosphoglycerol and phosphoethanolamine derived from membrane phospholipids and by succinyl residues. In this study, we describe the isolation of the opgE gene encoding the phosphoethanolamine transferase by a screen previously used for the isolation of the opgB gene encoding the phosphoglycerol transferase. Both genes show structural and functional similarities without sequence similarity.
Osmoregulated periplasmic glucans (OPGs), formerly membrane derived oligosaccharides are oligosaccharides accumulated in the envelope of many Gram-negative bacteria in media of low osmolarity. They belong to the common virulence factors found in phyto- and zoo pathogen of many Gram-negative bacterial species. In addition, OPGs control motility and secretion of exopolysaccharides in several species [7–9]. Glucose is the sole constitutive sugar and this glucosidic backbone may be substituted by various substituents depending on the species . In Escherichia coli, OPG backbone is synthesized by the opgGH products (Figure 1) . Total loss of OPGs reduced motility and increased exopolysaccharides synthesis and they are phenotypes associated with inactivation of opgG or opgH genes. OpgH is a transmembrane glucosyl-transferase catalyzing, with acyl carrier protein (ACP) as a cofactor and UDP-glucose as a substrate, the synthesis of a backbone constituted of linear β-1,2-linked glucose units. OpgG is a periplasmic glucosyl transferase branching glucose units on this linear backbone by β-1,6-linkages . This backbone is highly substituted by phosphoglycerol residues, and to a less extent by phosphoethanolamine residues, derived from membrane phosphatidylglycerol and phosphatidylethanolamine, respectively, and by succinyl residues (Figure 1) . Two phosphoglycerol transferases catalyze the substitution by phosphoglycerol and are encoded by the same opgB gene . The phosphoglycerol transferase I is anchored to the inner membrane but a large periplasmic catalytic domain transfers phosphoglycerol residues from phosphatidylglycerols to OPG molecules. The phosphoglycerol transferase II is a periplasmic soluble enzyme resulting from the liberation of the periplasmic catalytic domain of the former one and catalyzes the transfer of phosphoglycerol residues from one OPG molecule to another. Both phosphoglycerol transferases can transfer phosphoglycerol to artificial β-glucoside acceptors such as arbutin leading to enhanced turnover of phosphatidylglycerol. This was the basis of the selection of the mdoB mutants. In a dgk strain grown in media of low osmolarity containing arbutin, accumulation of diacylglycerol occurred in membrane to a toxic level and cell growth slowed abruptly. In a double mutant dgk opgB, diacylglycerol accumulation decreased and growth resumed . The succinyl-transferase is encoded by the opgC gene and is an intrinsic transmembrane protein catalyzing the transfer of succinyl residues probably from succinyl-CoA to OPG molecules. In an opgB strain, anionic character of OPGs is only provided by succinyl residues since phosphoethanolamine is a neutral substituent. This was the basis of the isolation of the opgC mutant which was isolated by screening an opgB strain using the severe difference of migration within a thin layer chromatography between neutral and anionic OPGs .
In this paper, we describe the isolation of a new gene, opgE, needed for substitution of OPGs by phosphoethanolamine in E. coli by using a similar selection protocol previously described for isolation of the opgB mutant.
2. Material and Methods
2.1. Bacterial Strains, Media, and Growth Conditions
E. coli strains and bacteriophage are listed in Table 1. Bacteria were grown with vigorous shaking at 37°C in Luria-Bertani broth (LB)  unless, otherwise, it is indicated. Solid media were obtained by adding agar at 15 g/liter. For motility measurement, highly wet substrate is required and agar was added to LB medium at 3 g/liter (swimming medium). 107 bacteria were spotted onto this swimming medium in a volume of 5 μL and the intensity of the halo of swimming was measured after 24 h of incubation at 37°C. When low osmolarity medium was required, LB without NaCl (LB-N) was added and used . For exopolysaccharides secretion, bacteria were streaked onto the SOB-Glycerol solid medium . Exopolysaccharides secretion level was measured by intensity of the slimy aspect of bacterial colonies (also called mucoidy) after 30 h of growth at 30°C. Antibiotics were used at the following concentrations: ampicillin and kanamycin at 50 μg/mL and chloramphenicol at 25 μg/mL. X-Gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) were used at a concentration of 40 μg/mL. Arbutin was used at a concentration of 90 mM.
2.2. Recombinant DNA Techniques
Standard procedures were performed for genomic and plasmid DNA extractions. gDNA purification was performed with the Nucleospin tissue (Macherey-Nagel). DNA purification was performed with the Nucleospin miniprep (Macherey-Nagel). Restriction enzymes (fermentas), T4 DNA ligase (fermentas), and Taq polymerase (fermentas) were used according to the manufacturer’s recommendations.
2.3. Transduction and Transformation
Transduction with phage P1vir and transformation by the rubidium chloride technique of E. coli were carried out according to Miller .
2.4. Tn5 Transposon Mutagenesis
Random Tn5 mutagenesis was performed with λNK467 (Table 1) according to standard procedures . Bacterial strain NFB775 was grown in LB liquid medium until mid log phase at 37°C in LB medium. Phage λNK467 was added to bacteria at a multiplicity of infection (MOI) of 1 and incubated without shaking for 30 min. to allow adsorbtion of phages and entry of phage DNA in bacterial cells. Bacteria (109) were then plated onto LB plates containing kanamycin and incubated for 48 h at 42°C. The kanamycin resistant clones obtained were only the result of transposition of Tn5 from λNK467 to the bacterial DNA since in these conditions, λNK467 became a suicide vector because it cannot induce neither lytic cycle (mutation in the replication genes Oam and Pam) nor lysogeny (thermosensitive mutation of the repressor CI857).
2.5. Tn5 Insertion Localization
Tn5 transposition insertions were localized on the E. coli chromosome by inverse PCR using Tn5CEcoRI and Tn5DHindIII primers (Table 1) after digestion of the chromosomal DNA with taqI.
2.6. Cloning of the opgE Gene and Construction of the opgE::cml Mutation
Plasmids and primers designed for PCR are listed in Table 1. A 2073 bp DNA fragment containing the opgE gene was amplified by PCR (foropgE and revopgE primers) from genomic DNA. For cloning of the opgE gene, this PCR product was digested by EcoRI and HindIII, cloned into pUC18Not and digested by the same enzymes and transformed into competent E. coli cells to give pNF752. For construction of the opgE::cml mutation, the chloramphenicol cassette used for gene inactivation was released from pNFCml  after digestion by EcoRV and was inserted into the unique HpaI restriction site of the 2073 bp PCR DNA fragment containing opgE. The resulting DNA fragment was introduced into competent E. coli cells containing the pKD46 plasmid allowing homologous recombination by expression of the red recombinase gene products . The accurate insertion of the mutation was controlled by PCR using genomic DNA of the resulting opgE::cml mutant as a template and upopgE and C1 and downopgE and C2 as primers.
2.7. Extraction of OPGs
Bacteria were grown overnight in LB without NaCl (200 mL). Bacteria were collected by centrifugation at 4°C for 15 min at 8,000 ×g. Cell pellets were resuspended in 20 mL of distilled water and extracted with 5% trichloroacetic acid (TCA). The TCA extracts were neutralized with ammonium hydroxide 10% and concentrated by rotary evaporation. The resulting material (2 mL) was then fractionated by gel filtration on a Bio-Gel P-4 column (Bio-Rad, 55 by 1.6 cm) equilibrated with 0.5% acetic acid. The column was eluted in the same buffer at a flow rate of 15 mL h−1, and fractions of 1.5 mL were collected. Presence of sugar in each fraction was determined colorimetrically by the anthrone reagent procedure. Fractions containing OPGs were pooled and total content was determined by the same procedure . For mass spectrometry analysis, OPGs were subsequently desalted in water on a Bio-Gel P-2 column (Bio-Rad, 90 by 1.6 cm) and fractions containing OPGs were lyophilized.
2.8. Mass Spectrometry
All mass spectra were acquired on a Voyager Elite (DE-STR) reflectron time-of-flight (TOF) mass spectrometer (Perseptive Biosystems, Framingham, MA), equipped with a pulsed nitrogen laser (337 nm) and a gridless delayed extraction ion source. Oligosaccharide samples were analyzed in delayed extraction mode using an accelerating voltage of 20 kV, a pulse delay time of 200 ns, and a grid voltage of 66%. Detector bias gating was used to reduce the ion current for masses below 500 Da. Between 100 and 200, scans were averaged for each mass spectrum. Oligosaccharide alditols were cocrystallized with 2,5-dihydroxybenzoic acid (DHB) as matrix (10 mg mL−1 of DHB in methanol/water (50/50) containing 0.1% trifluoro acetic acid (TFA)). For all measurements, the “dried droplet” preparation technique was used. Typically, 1 mL of the matrix was mixed on-target with 1 μL of water-dissolved oligosaccharides and allowed to dry under an air stream. They were analyzed in positive ion mode.
3. Results and Discussion
3.1. Arbutin Is a Substrate for the Phosphoethanolamine Transferase
Arbutin is an artificial substrate for the phosphoglycerol transferase and may be also substrate of the phosphoethanolamine transferase since they catalyze similar reactions (transfer of polar head from phospholipids to OPGs). In a dgk strain, arbutin resistance was used to isolate the opgB gene (see Section 1) and similar procedure could be used to isolate mutant devoid of the phosphoethanolamine transferase. The screen could be used only if severe differential growth of bacterial colonies was shown between the NFB775 grown without arbutin (normal growth) or with arbutin (slow growth resulting from accumulation of toxic diacylglycerol) (Figure 2). The NFB775 (opgB dgk) strain, synthesizing OPGs substituted with succinyl and phosphoethanolamine residues but devoid of phosphoglycerol residues, was grown at 37°C in LB-N liquid medium overnight. Appropriate dilution of the culture was performed and the same volume of dilution was spotted onto 6 LB-N medium plates and onto 6 LB-N medium plates with Arbutin 90 mM being added. These 12 plates were incubated at 37°C and colony forming units (CFU) were numbered in the same 12 plates after 24 h, 48 h, and 72 h of incubation. In LB-N medium, an average of 243 CFU of normal size were numbered after 24 h of incubation and 260 CFU were counted after 48 h or 72 h of incubation. In LB-N medium with arbutin, an average of 15 very small CFU were hardly detected after 24 h of incubation, 146 small CFU after 48 h of incubation, and 207 medium CFU after 72 h of incubation. Thus, growth is severely impaired in NFB775 grown in LB-N with arbutin added as compared to the same medium without arbutin added and colony size were still smaller after 72 h of growth in medium containing arbutin as compared to colonies grown 24 h in medium without arbutin added. Thus, isolation of arbutin resistant mutants could be performed. Because the dgk strain tested was devoid of phosphoglycerol transferase (opgB), one can imagine that arbutin is also a substrate for phosphoethanolamine transferase stimulating its activity, increasing phosphatidylethanolamine turnover leading to accumulation of diacylglycerol toxic for bacteria (Figure 2) and that at least one kind of arbutin resistant mutants could result from the inactivation of the gene encoding the phosphoethanolamine transferase.
3.2. Screening of an opgE::Tn5 Mutant
Random transposon mutagenesis with Tn5 was performed on the NFB775 strain (see Section 2). 30000 clones kanamycin resistant were collected and divided into 20 pools of 1250 clones. One LB-N plate containing 90 mM arbutin was used to plate independently each of the 20 pools and incubated at 37°C. The first 30 clones that appeared on these plates (average of 1.5 clone per plate) were collected after 24 h, purified, and grown in LB-N liquid medium. One LB-N plate containing 90 mM arbutin was used to plate independently 200 CFU of each of the 30 cultures and incubated at 37°C. The 5 plates harboring bacterial colonies growing faster than the others were selected and one clone isolated from each of the 5 plates was further studied. The 5 genes disrupted by transposon were located on the E. coli chromosome (see Section 2). The first two transposons were inserted in the intergenic DNA region lying between the malK and lamB genes of the malK lamB malM operon of the maltose regulon, thus inactivating lamB and MalM. The third transposon was inserted within the yidL gene encoding a putative AraC regulatory protein. The fourth transposon was inserted within fucI encoding a L-fucose isomerase. The fifth transposon was inserted within ybiP encoding a putative inner membrane protein with a sulfuric ester hydrolase and transferase activities domains (accession number Ecocyc G6418). These two domains were also found on the OpgB protein (accession number Ecocyc EG12591). Thus, we decided to further characterize the ybiP mutant clone.
3.3. The opgE Gene Encodes the Phosphoethanolamine Transferase
Mass Spectrum of OPGs extracted from the ybiP mutant clone grown in LB-N medium was performed. An increment of 100 Da, corresponding to substitution of OPGs by succinyl residues was observed for several molecular ions. No increment of 123 Da, corresponding to phosphoethanolamine substitution of OPGs, was observed (Figure 3). This strongly suggests that ybiP, renamed opgE, encodes the phosphoethanolamine transferase. To confirm that opgE encodes the phosphoethanolamine transferase, opgE was cloned into pUC18Not (pNF752) and inactivated on the E. coli chromosome by reverse genetic (opgE::cml) (see Section 2). Mass spectra of OPGs extracted from NFB2228 (opgB opgC), NFB4548 (opgB opgC opgE), and NFB2231 (NFB4548/pNF752) were compared (Figure 3). The sizes of OPGs varied from 6 to 12 residues of glucose, as illustrated on MS spectra by the [M+Na]+ molecular ions at m/z 1013, 1175, 1337, 1499, 1661, 1823, and 1985, respectively. As expected, OPGs were neither substituted by succinyl residues nor by phosphoglycerol residues in these mutants. OPGs extracted from NFB4548 strain, in which opgE has been inactivated, were not substituted by phosphoethanolamine residues (Figure 3). On the contrary, in NFB2228 and NFB2231 strains, in which opgE is functional, a mass increment of 123 Da was observed for OPGs, with, for example, molecular ions at m/z 1298, 1460, 1622, 1784, and 1946 (Figure 3). This mass increment corresponded to the substitution of OPGs by phosphoethanolamine residues and confirmed that opgE encodes the phosphoethanolamine transferase. Taken together, these results suggest that arbutin is not only a substrate for the phosphoglycerol transferase but also for the phosphoethanolamine transferase (Figure 2). Nevertheless, the arbutin resistance is probably a more complex phenomenon than previously thought. The reason for the arbutin resistance of the 4 other mutants is difficult to explain and is probably not related to OPGs suggesting that arbutin affects several different cellular processes.
(a) dgk opgB opgE (ybiP)
(b) opgB opgC
(c) opgB opgC opgE
(d) opgB opgC opgE/pl opgE
3.4. Sequences Analysis of opgE
opgE is an operon constituted of one gene transcribed counterclockwise. The 1584 bp open reading frame begins at 851 820 bp and ends at 850 237 bp on the E. coli K12 chromosome and encodes a 527 amino acid protein (Figure 4) of 59 707 Da. The TMHMM2.0 algorithm  allowed the prediction of four transmembrane spanning segments in the 166 first amino acids and a periplasmic domain in the 361 following amino acids. The 217 to 483 amino acid sequence form a 226 amino acids catalytic domain with sulfuric ester hydrolase and transferase activities. Arbutin cannot cross the inner membrane of E. coli K12 strongly suggesting a periplasmic location of this catalytic domain. The OpgB protein shows a very similar organization  with 3 or 4 transmembrane segments in the 162 first amino acids followed by a periplasmic domain containing a 285 amino acids sulfuric ester hydrolase and transferase activities (amino acids 163 to 448) . Despite the similarity between structure and function (except the donor phospholipid) of these two enzymes, no significant similarity was observed between the OpgB and OpgE amino acids sequences even in the hydrolase and transferase catalytic domains.
3.5. Phenotypes of E. coli Strains Devoid of Substituents
The main phenotypes of mutant strains devoid of OPGs backbone (i.e., opgG or opgH) are loss of motility and increased exopolysaccharides synthesis (see Section 1). We decided to test these phenotypes in strains devoid of one or all kinds of substitution on OPGs backbone. For motility measurement, opgB (NFB732), opgC (NFB1887), opgE (NFB4576), and opgB opgC opgE (NFB4548) strains were spotted on swimming agar (see Section 2) for 24 h together with the wild type (JM83) and the opgH (NFB216) strains as controls. As expected for the two controls, the wild-type strain was motile while motility was severely reduced for the opgH strain. As observed for the wild-type strain, all the mutant strains tested were motile (Figure 5). For exopolysaccharides synthesis measurement, the same strains were plated onto SOB-Glycerol medium. As observed for the wild-type strain, all the mutant strains were non mucoid indicating no increase in exopolysaccharides synthesis except for the opgH mutant strain where mucoidy was increased indicating an increase in exopolysaccharides synthesis (Figure 5). Thus, no classical phenotype found in mutants devoid of OPGs backbone could be associated with mutants only devoid of substituent. Thus, the role of substituent remains to be elucidated.
The isolation of the opgE gene strongly suggests that the minimum set of genes required for a complete biosynthesis pathway of OPGs were characterized (Figure 1). The suggested interactions  between several of these enzymes suggest that these enzymes work within a complex which remains to be established. The role and the reason of the difference of substitution depending on the species remain to be established but the absence of phenotype renders this analysis difficult.
Sébastien Bontemps-Gallo and Virginie Cogez contributed equally to this work.
The authors thank Laury Jarworski for technical assistance. This work was supported by grants from the Centre National de la Recherche Scientifique and from the Ministère de l’Education Nationale et de la Recherche.
J.-M. Lacroix, M. Tempête, B. Menichi, and J.-P. Bohin, “Molecular cloning and expression of a locus (mdoA) implicated in the biosynthesis of membrane-derived oligosaccharides in Escherichia coli,” Molecular Microbiology, vol. 3, no. 9, pp. 1173–1182, 1989.View at: Google Scholar
J.-M. Lacroix, E. Lanfroy, V. Cogez, Y. Lequette, A. Bohin, and J.-P. Bohin, “The mdoC gene of Escherichia coli encodes a membrane protein that is required for succinylation of osmoregulated periplasmic glucans,” Journal of Bacteriology, vol. 181, no. 12, pp. 3626–3631, 1999.View at: Google Scholar
F. Bouchart, G. Boussemart, A. F. Prouvost et al., “The virulence of a Dickeya dadantii 3937 mutant devoid of osmoregulated periplasmic glucans (OPGs) is restored by inactivation of the RcsCD RcsB phosphorelay,” Journal of Bacteriology, vol. 192, no. 13, pp. 3484–3490, 2010.View at: Publisher Site | Google Scholar
W. Ebel, G. J. Vaughn, H. K. Peters III, and J. E. Trempy, “Inactivation of mdoH leads to increased expression of colanic acid capsular polysaccharide in Escherichia coli,” Journal of Bacteriology, vol. 179, no. 21, pp. 6858–6861, 1997.View at: Google Scholar
S. Bontemps-Gallo, E. Madec, J. Dondeyne et al., “Concentration of osmoregulated periplasmic glucans (OPGs) modulates the activation level of the RcsCD RcsB phosphorelay in the phytopathogen bacteria Dickeya dadantii,” Environmental Microbiology, vol. 15, pp. 881–894, 2013.View at: Google Scholar
J. P. Bohin and J. M. Lacroix, “Osmoregulation in the periplasm,” in The Periplasm, M. Ehrmann, Ed., pp. 325–341, American Society for Microbiology, Washington, DC, USA, 2006.View at: Google Scholar
J.-M. Lacroix, I. Loubens, M. Tempete, B. Menichi, and J.-P. Bohin, “The mdoA locus of Escherichia coli consists of an operon under osmotic control,” Molecular Microbiology, vol. 5, no. 7, pp. 1745–1753, 1991.View at: Google Scholar
Y. Lequette, E. Lanfroy, V. Cogez, J.-P. Bohin, and J.-M. Lacroix, “Biosynthesis of osmoregulated periplasmic glucans in Escherichia coli: the membrane-bound and the soluble periplasmic phosphoglycerol transferases are encoded by the same gene,” Microbiology, vol. 154, no. 2, pp. 476–483, 2008.View at: Publisher Site | Google Scholar
B. J. Jackson, J.-P. Bohin, and E. P. Kennedy, “Biosynthesis of membrane-derived oligosaccharides: characterization of mdoB mutants defective in phosphoglycerol transferase I activity,” Journal of Bacteriology, vol. 160, no. 3, pp. 976–981, 1984.View at: Google Scholar
J. H. Miller, A Short Course in Bacterial Genetics: A Laboratory Manual and Handbook for Escherichia coli and Related Bacteria, Cold Spring Harbor Laboratory Press, New York, NY, USA, 1992.
T. Silhavy, M. Berman, and L. Enquist, Experiments with Gene Fusions, Cold Spring Harbor Laboratory Press, New York, NY, USA, 1984.
E. L. Sonnhammer, G. von Heijne, and A. Krogh, “A hidden Markov model for predicting transmembrane helices in protein sequences,” Proceedings of the International Conference on Intelligent Systems for Molecular Biology, vol. 6, pp. 175–182, 1998.View at: Google Scholar